Taking advantage of a little bit of good sky, I'm doing some tests from home with real tools.
This Plate Solver works but only partially.
If the object is in the center enough, it is solved and everything is ok.
But if the object is not really in the center, or if I go on a
a different object and this too is not near the center,
he makes a first correction and if he feels, because the mount moves.
But in the second step, even if the object has not yet been resolved,
everything stays there! In fact, you can see the various points that always remain
in the same place, so the mount is moved only once
the first step and not other times.
I hope I explained myself.

I would not recommend the "slew to target" option. I would recommend plate solving using the "sync" option. If you need to recenter it after the plate solve, do that separately. I think you will have more luck with that. If you plate solve too much in a small area, that can throw off your pointing model with small mount pointing errors.

I would definitely recommend first doing several plate solves in different parts of the sky using the sync option, do NOT do slew to target. Then the mount can build its model. When you are getting more accurate slews, then try to go to your object. If it is off, plate solve nearby and then try again. But don't just keep solving and solving.

Currently imaging using 3.4.0. EQ Mod mount, ZWO Guidescope, Canon 5D Mk ii and Moonlite focuser. Moonlite focuser serial timeout error still there, but I can manually focus. Plate solver (offline) works perfectly. Good guiding and Canon imaging away.

I'm sorry and a little ashamed,
but the translation was not good e
I'm confused.
I wanted to point out that I'm using a Vixen SXP mount,
an 80/352 refractor and a modified Canon Eos 60D.

Incidentally I ran 3.4 two nights in a row on my laptop. Schedule, sequence, alignment, focus, guiding, and PHD2 all check out. No crashes or strange errors.

Lead_weight and cxxone,

That sounds good! If there are any issues, please let me know. I think Jasem wants to make this release a good stable release, so any bugs or kinks should be worked out.

Lead_weight,

Later this week, maybe I will experiment with Python in a virtual OS X machine to see if I can replicate the situation your desktop is having and see if I can work around it. That might be the best option instead of trying to figure it out remotely. Im glad the laptop is working fine.

carkinzo,

If it wasn't clear, maybe a screenshot would help. I will also try to explain again more clearly.

Do not do plate solve with "Slew to Target", do plate solve with "Sync"


When you first set up your telescope, do several plate solves in different parts of the sky. Don't worry if the target does not show up in the image when you do a plate solve. It does not matter. The mount is still getting told where the telescope is pointing, even if the target isn't in the picture. Each plate solve improves the accuracy of the "mount model," assuming that your mount can build a model (some can and some cannot). If your mount cannot build a pointing model, it will still be good to use at least 3 alignment points by doing plate solves in different parts of the sky. Then after you do a few plate solves, the accuracy should improve. Finally do a plate solve near where you actually plan to be imaging.

In my opinion it is not a good idea to do "Slew to Target" because then you get several alignment points that are very close together. You really want your alignment points to be far apart to produce a better model of the sky.

I agree with you in headlines. But I myself usually do not image more than 1 or 2 objects in one night and in that case "Slew To Target" works really well. My SkyWatcher HEQ5 with EQMod always puts the object that I want to image within 30 arcsec of the center of FOV within 5 or 6 tries. All I mean to say is: I think it depends both on your mount and on your purpose which option to choose.


Wouter

Sounds, good. PM me if you think a quick screen share session to poke around would be helpful, and we can set up a time.

Yes, this is certainly true, and that is why the option exists to do "slew to target." However, he said he was having problems due to the mount not slewing to objects properly after plate solves. There are several things that can cause this that I can think of (pointing model messed up by points too close together, mounts not very accurate with short slews, etc), and most of them would be solved by following the advice that I gave. I find that my own mount struggles when doing plate solves with "slew to target," but. that it works much better if I do several plate solves that are widely spaced apart to build a pointing model before I go to my object. And once I do, that, my slews are pretty much spot on. So yes, certainly it depends on your situation, but I think in this situation, it might help to try my alternate approach.

Yes you are absolutely right. I didn’t mean to contradict you Rob because you give very good advice. Let’s see if carkinzo can solve his slewing option by using the sync option and by building a pointing model first.

The thing that seemed strange to me is that the first correction
does it, but then no longer corrects!
Okay, I'll have to test, but the sky is here again
cloudy, so I don't know when I can do it.
I wanted to ask one thing:
I had heard that to work better you should do it
a 1 or 2 star alignment with the mount computer.
Is it right to do this, or is it useless?
I ask because my Vixen Sxp already with a 3 alignment
stars is accurate, so I asked for this reason.
But if this were right, I would have to sync the map
the StarBook Ten with Kstars and I don't know if it can be done.